A plant monitor is a small circuit that watches three things: how wet the soil is, how much light the plant gets, and how warm the air is. When a reading drifts out of range, it raises an alert. Build one and you have covered analog sensors, calibration, threshold logic and a weatherproof enclosure in a single project. This walkthrough takes a classroom build from bare parts to a working monitor, and finishes with the part that matters most in a South African school: keeping the plant alive through the holidays.

What a plant monitor actually does

Strip away the enclosure and the code and the job is simple. A microcontroller reads a value from each sensor a few times a minute, compares each value against a threshold you set, and acts when a threshold is crossed. Acting can mean lighting an LED, sounding a short beep, showing a message on a screen, or later switching a small pump. Everything else is detail on top of that loop: read, compare, act, wait, repeat.

The three signals each answer a different question. Soil moisture tells you whether to water. Light tells you whether the plant is in the right spot. Temperature tells you whether the room is safe for it, which matters in a classroom that bakes in summer and turns cold on a winter morning.

The parts you need

Keep the parts list short and cheap so a class set stays affordable. You need one of everything below per student, or per pair for younger groups.

• A microcontroller board with analog inputs. The sheenbot∞ board handles this project comfortably, and its block editor keeps younger classes moving.
• A soil moisture sensor. Prefer a capacitive one; the cheap resistive probes corrode in wet soil within weeks.
• A light sensor, such as an LDR, with a resistor to form a voltage divider.
• A temperature sensor, either an analog thermistor or a small digital sensor.
• A breadboard, jumper wires and a way to power the board.
• An output for alerts: an LED and resistor, a buzzer, or the board's screen.

Add a few spare sensors to the order; probes are the first thing to fail. Budgeting roughly ten to fifteen percent of the kit cost for spares saves a scramble mid-term. Full kits and individual parts are in the store.

Wiring and the first sensing loop

Wire one sensor at a time and test it before adding the next. Connect power and ground, run the sensor's signal line to an analog input, and write a short loop that reads the value and prints it. You are not looking for correct numbers yet, only for numbers that change when you change the world: press the moisture probe into a damp sponge and the reading should move, cover the light sensor with your hand and it should drop.

Once each sensor moves in the right direction, print all three together on one line every few seconds. This raw stream is the foundation for the next step, and it is also the first place students see that a sensor never returns a clean, round number.

Calibration is the science lesson

A raw moisture reading means nothing on its own. Turning it into dry or wet is where the real learning sits, and it is a proper experiment. Take two known states: bone-dry soil and fully soaked soil. Record the sensor value for each, several times, and average them. Those two numbers are your endpoints, and any reading in between maps onto a percentage from dry to wet.

Run this as a controlled investigation. Change one thing at a time, keep the probe at the same depth, and write every reading in a table before you touch the code. Students quickly notice that the endpoints drift between soil types and between a warm and a cold room, which is exactly the point: calibration is not a one-off, and no sensor is absolute. Repeat the same two-point method for light, a dark drawer against a bright window, and for temperature against a classroom thermometer.

Thresholds and alert logic

With calibrated readings you can set thresholds that mean something. Decide, per plant, the moisture percentage below which it needs water, a light level that flags a bad spot, and a temperature band it should stay inside. The logic is then a set of plain checks:

1. If moisture is below the dry threshold, raise the water-me alert.
2. If light stays below its threshold for a while, flag the position.
3. If temperature leaves the safe band, warn about heat or cold.

Add a small margin, or hysteresis, so an alert does not flicker on and off around the threshold. A common fix is to require the reading to sit past the line for several checks in a row before the alert fires, and to clear it only once the reading has comfortably recovered. This one idea prevents most of the beeping-at-random frustration.

The enclosure

Water and electronics do not mix, so the build is only finished when it is housed. Keep the board, battery and any exposed wiring in a closed box, and let only the sensor probes reach out to where they measure. Bring the moisture probe out low near the soil, the light sensor out to a window-facing side, and leave a vent gap so trapped heat does not skew the temperature reading. A clear lid lets students see the LED without opening anything, and recycled containers work well while keeping the cost down.

Keeping the plant alive over the holidays

The monitor is most useful exactly when nobody is in the room. Three extensions turn it into something that survives a school break. First, log readings to the board's storage with a timestamp, so the class can see what happened while they were away. Second, add a small water pump triggered by the moisture alert so the plant waters itself, with a hard limit on pump time so a stuck sensor cannot flood the pot. Third, plan for power: mains fails during load shedding, so a battery with a generous margin, or a low-power sleep between readings, keeps the monitor running through an outage.

This holiday-proofing is a natural project for a school break itself. Our holiday workshops run builds like this from bare parts to a finished enclosure, which is a good format for the extension work that a normal lesson slot rarely has time for.

Takeaway

A plant monitor is a rare project that is genuinely useful, cheap to build, and rich enough to carry real science. Build the sensing loop first, treat calibration as the experiment it is, and the alert logic and enclosure fall into place. Do the holiday extension and both the plant and the lesson survive the break.

Common questions

Which sensor should we buy first if the budget is tight?

The soil moisture sensor, capacitive type. Moisture is the reading students act on most, the one that drives watering, and it is the sensor most in need of the calibration lesson. Light and temperature can be added in a later session.

Why does the moisture reading jump around even in still soil?

Small electrical noise is normal. Average several readings before comparing against a threshold, and keep the probe at a fixed depth. If the numbers swing wildly, check the connections and the power line before you blame the sensor.

What age group is this project for?

It scales. Upper primary can do it with block code and a guided calibration table, while high school can write the threshold logic themselves and analyse the logged data. The academy pitches the same build at different levels.